en.wikipedia.org/wiki/Biological_soil_crust#/media/File:Cryptobiotic.jpg

This sentence defines biocrusts as if they occur only in arid and semi-arid ecosystems. That is too narrow and factually misleading.

Authoritative reviews describe biocrusts as common in drylands but also present in mesic/temperate settings. The U.S. Forest Service review says biocrusts are common in arid and semiarid regions, but are also present in more mesic areas such as the Great Plains. A separate review focused specifically on temperate systems describes them as "mesic counterparts" and says biocrusts in temperate biomes are often widespread.

So the problem is not that drylands are unimportant — they are the best-studied and often the most extensive biocrust habitats — but that the article's definition wrongly restricts biocrusts to arid and semi-arid ecosystems.

The numerical range here is familiar in the biocrust literature, but the unit has been mangled.

Published sources report annual biocrust carbon inputs as 0.4 to 37 g C m⁻² yr⁻¹ (grams of carbon per square meter per year), not g/cm*year. The Wikipedia sentence therefore misstates the unit, which materially changes the meaning of the estimate.

This looks like a unit-formatting or transcription error rather than a disagreement in the literature, but it is still a factual error because the quantity as written is not the one reported by the cited research.